skinny.

Latest / Taylor Tailored

Can Humans Keep Control of Advanced AI?

AI leaders have warned that increasingly powerful systems might eventually become difficult for people to supervise or stop. During a week of high-level UN discussions, the argument raises practical questions that go beyond dramatic predictions.

What is an AI agent allowed to do? How can evaluators test a system that uses tools and takes several steps towards a goal? What does “misalignment” mean, and how could governments check safety commitments made by developers?

This Taylor Tailored episode separates capabilities seen in present systems from uncertain forecasts about rapid…

Log in to save or queue this episode.

The skinny

The skinny isn't ready yet — notes appear once the transcript is processed.